mlsanityData to 7 Oct 2026

On the Limitation and Experience Replay for GNNs in Continual Learning

arXiv 2302.03534 ↗First posted 7 Feb 2023cs.LGcited by 7

Not on the map. Why?

Abstract (arXiv)

Continual learning seeks to empower models to progressively acquire information from a sequence of tasks. This approach is crucial for many real-world systems, which are dynamic and evolve over time. Recent research has witnessed a surge in the exploration of Graph Neural Networks (GNN) in Node-wise Graph Continual Learning (NGCL), a practical yet challenging paradigm involving the continual training of a GNN on node-related tasks. Despite recent advancements in continual learning strategies for GNNs in NGCL, a thorough theoretical understanding, especially regarding its learnability, is lacking. Learnability concerns the existence of a learning algorithm that can produce a good candidate model from the hypothesis/weight space, which is crucial for model selection in NGCL development. This paper introduces the first theoretical exploration of the learnability of GNN in NGCL, revealing that learnability is heavily influenced by structural shifts due to the interconnected nature of graph data. Specifically, GNNs may not be viable for NGCL under significant structural changes, emphasizing the need to manage structural shifts. To mitigate the impact of structural shifts, we propose a novel experience replay method termed Structure-Evolution-Aware Experience Replay (SEA-ER). SEA-ER features an innovative experience selection strategy that capitalizes on the topological awareness of GNNs, alongside a unique replay strategy that employs structural alignment to effectively counter catastrophic forgetting and diminish the impact of structural shifts on GNNs in NGCL. Our extensive experiments validate our theoretical insights and the effectiveness of SEA-ER.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 40 of its 62 references that are in the map. First posted on arXiv 7 Feb 2023.

Builds on

40 of 62 references.

References · 40

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  2. 2017Inductive Representation Learning on Large Graphscited by 6,284
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2019A Comprehensive Survey on Graph Neural Networkscited by 2,906
  5. 2016Learning without Forgettingcited by 2,465
  6. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,148
  7. 2017Gradient Episodic Memory for Continual Learningcited by 1,871
  8. 2016Progressive Neural Networkscited by 1,332
  9. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
  10. 2017Continual Learning with Deep Generative Replaycited by 1,152
  11. 2019Large Scale Incremental Learningcited by 727
  12. 2017Lifelong Learning with Dynamically Expandable Networkscited by 689
28 more
  1. 2019Gradient based sample selection for online continual learningcited by 568
  2. 2019Orthogonal Gradient Descent for Continual Learningcited by 361
  3. 2021Gradient Projection Memory for Continual Learningcited by 338
  4. 2017Deep Gaussian Embedding of Graphs: Unsupervised Inductive Learning via Rankingcited by 332
  5. 2019Representation Learning for Dynamic Graphs: A Surveycited by 260
  6. 2020Supermasks in Superpositioncited by 223
  7. 2020Continual Lifelong Learning in Natural Language Processing: A Surveycited by 152
  8. 2018Contextual Stochastic Block Modelscited by 134
  9. 2019Scalable and Order-robust Continual Learning with Additive Parameter Decompositioncited by 125
  10. 2020A survey on domain adaptation theory: learning bounds and theoretical guaranteescited by 109
  11. 2016Less-forgetting Learning in Deep Neural Networkscited by 97
  12. 2020Overcoming Catastrophic Forgetting in Graph Neural Networkscited by 93
  13. 2018Streaming Graph Neural Networkscited by 91
  14. 2020Optimal Continual Learning has Perfect Memory and is NP-hardcited by 86
  15. 2021Subgroup Generalization and Fairness of Graph Neural Networkscited by 69
  16. 2020Streaming Graph Neural Networks via Continual Learningcited by 63
  17. 2020Graph Neural Networks with Continual Learning for Fake News Detection from Social Mediacited by 43
  18. 2020GraphSAIL: Graph Structure Aware Incremental Learning for Recommender Systemscited by 42
  19. 2021Graph Few-shot Class-incremental Learningcited by 40
  20. 2021Continual Learning of Knowledge Graph Embeddingscited by 35
  21. 2020Lifelong Graph Learningcited by 31
  22. 2022Graph Lifelong Learning: A Surveycited by 30
  23. 2021Hierarchical Prototype Networks for Continual Graph Representation Learningcited by 28
  24. 2020Disentangle-based Continual Graph Representation Learningcited by 25
  25. 2020Lifelong Learning of Graph Neural Networks for Open-World Node Classificationcited by 22
  26. 2023Continual Graph Learning: A Surveycited by 17
  27. 2022A Theory for Knowledge Transfer in Continual Learningcited by 16
  28. 2021Catastrophic Forgetting in Deep Graph Networks: an Introductory Benchmark for Graph Classificationcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 1; 2026: 2.

7 papers.

  1. 2024Continual Learning on Graphs: Challenges, Solutions, and Opportunitiescited by 20
  2. 2024Towards Robust Graph Incremental Learning on Evolving Graphscited by 18
  3. 2024Towards Continuous Reuse of Graph Models via Holistic Memory Diversificationcited by 7
  4. 2024Temporal-Aware Evaluation and Learning for Temporal Graph Neural Networkscited by 5
  5. 2026Structuring Semantic Embeddings for Principle Evaluation: A Prototype-Guided Contrastive Learning Approachcited by 1
  6. 2025The Unreasonable Effectiveness of Randomized Representations in Online Continual Graph Learningcited by 0
  7. 2026Structuring Relations Among Learning Paradigms via Protocol--Objective--Resource Reductionscited by 0